whereismyai
VisionClaw
VisionClaw is a real-time AI assistant for Meta Ray-Ban smart glasses that combines voice and vision via Gemini Live.
VisionClaw provides a real-time voice and vision AI experience for Meta Ray-Ban smart glasses and mobile phones. It streams camera input and bidirectional audio to the Gemini Live API, allowing users to talk to the AI about their surroundings. An optional OpenClaw gateway routes tasks to connected apps, enabling actions like web searches and messaging. The software includes native source code implementations for both iOS and Android platforms, utilizing the respective Meta Wearables DAT SDKs. Developers can self-host the gateway, configure custom API keys, and test functionality using a phone camera mode.
VisionClaw at a glance
- Best for: Software engineer
- Runs on: iOS, Android, Web, self-hosted, open source
- Pricing: Free
- Released: February 2026
- Category: AI agent tool
- Website: github.com/Intent-Lab/VisionClaw
Key features
- Gemini Live real-time voice and vision AI over WebSocket
- Meta Ray-Ban smart glasses camera and audio integration
- Optional OpenClaw gateway for 56+ tools and connected apps
- Phone mode for testing pipelines using the device camera
- WebRTC streaming to share the camera point of view live
- Configurable audio pipelines for echo cancellation and mic gating
- Function calling that routes user tasks to external tools
Use it for
- Describe the physical scene seen through smart glasses
- Add items to shopping lists using voice commands
- Send messages via WhatsApp, Telegram, or iMessage
- Search the web for local spots and places nearby
Companions
Tools that articles about VisionClaw recommend using with it, each with its job.
- OpenClaw — Built to work with OpenClaw.
- Composio — Equips AI agents with tools, managed user auth, and sandboxes across 1,500+ apps
- Supermemory — Integrates long-term memory into AI coding agents with a single prompt.
- memU — Hierarchical knowledge graph for proactive behaviour.
- MemOS — Shared memory across agents, also available as a cloud service.
Alternatives to VisionClaw
- PhoneClaw — It executes AI agent tasks and system tool calls entirely on-device.
- Vision Agents — It enables building low-latency AI agents with support for any model.
- MetaClaw — Continuously improves your personal agent from live chats without requiring a local GPU cluster.
- NemoClaw — Provides managed security, networking, and lifecycle operations for AI agents.
- NanoClaw — Runs in isolated Docker containers with credential protection via OneCLI.
- ZeroClaw — Own your AI infrastructure with no subscriptions or cloud dependencies.
- SkillClaw — Automatically evolves and unifies agent skills across devices and team members.
- thClaws — Runs locally in native Rust with zero cloud lock-in and multi-provider support.
- GoClaw — It offers Go-native performance with multi-tenant isolation and 5-layer security.
- MateClaw — It secures data and workflows within a self-hosted, audit-ready environment.
- PokeClaw — Operates directly on-device without needing a cloud connection or API keys.
- Qwen MM Plugins — Enables agents to process multimodal data natively rather than via APIs.
All VisionClaw alternatives